skills.rest
SciMate-AI
SciMate-AIOFFICIAL
Explore all skills by this author →
GitHub

diffdock

Predict protein-ligand binding poses and confidence scores from PDB files or sequences with SMILES inputs.

1|Updated Mar 11, 2026
Search Tags:#drug discovery#molecular docking#protein-ligand#diffusion-models#pose prediction#diffdock
One-click install
npx skills add https://github.com/SciMate-AI/scicli --skill diffdock-scimate-ai
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: diffdock
Source: https://github.com/SciMate-AI/scicli/tree/main/internal/skills/bundled/claude-scientific-skills/skills/diffdock
Command: npx skills add https://github.com/SciMate-AI/scicli --skill diffdock-scimate-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pandas, rdkit, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

DiffDock provides end-to-end docking workflow to predict three-dimensional binding poses for ligand molecules relative to protein targets, along with per-pose confidence scores, enabling rapid structure-based drug design.

Core Features & Use Cases

  • Single-protein-ligand docking
  • Batch docking and virtual screening
  • ESMFold sequence support
  • Pose ranking and confidence scoring
  • Downstream rescoring with GNINA/MM-PBSA

Quick Start

Provide a protein structure (PDB) or sequence and a ligand (SMILES or structure), then run the configured inference to generate ranked docking poses with confidence scores.

Frequently Asked Questions about diffdock

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I predict protein-ligand binding poses from PDB files and SMILES strings?▼

Predict protein-ligand binding poses by providing a protein structure (PDB) or sequence and a ligand SMILES string to run diffusion-based sampling, which outputs ranked 3D docking poses (SDF) with confidence scores.

Can I use a protein sequence for molecular docking instead of a PDB structure?▼

Yes, molecular docking supports ESMFold sequence inputs, allowing you to predict binding poses directly from protein sequences paired with ligand SMILES without requiring a pre-existing PDB structure.

How do I run batch docking for virtual screening?▼

Batch docking for virtual screening is executed through a YAML-driven inference workflow, processing multiple protein-ligand pairs to generate ranked poses and per-pose confidence scores for rapid structure-based drug design.

What is the best way to rank and score predicted docking poses?▼

Rank and score predicted docking poses using built-in diffusion-based confidence scoring, with optional downstream rescoring available through GNINA or MM-PBSA for extended virtual screening workflows.

Do I need rdkit and pandas to run diffusion-based pose prediction?▼

Yes, diffusion-based pose prediction requires rdkit and pandas dependencies to process molecular inputs, handle ligand structures, and manage the data outputs for the docking workflow.

Are there limitations to single protein-ligand docking with diffusion models?▼

Single protein-ligand docking using diffusion models is optimized for rapid structure-based drug design, but complex systems may require downstream rescoring with GNINA or MM-PBSA to validate binding pose accuracy.

Related Skills

Explore More Education & Research Skills→
swaruplabswaruplab
94

diffdock

Predict protein-ligand binding poses and confidence scores from PDB and SMILES inputs.

Community
Advanced
LeonChaoXLeonChaoX
783

diffdock

Predict ranked 3D protein–ligand binding poses with confidence scores.

Community
Advanced
Estrella-231Estrella-231

diffdock

Predict ranked protein–ligand binding poses from PDB or SMILES inputs.

Community
Intermediate
jimmc414jimmc414
557

diffdock

Predicts protein-ligand binding poses and confidence scores via molecular docking.

Community
Advanced